I recently heard the following story. It begins in 1939 when academic Eliot D. Smith wrote a study about textile mills in the United States during the Great Depression. These mills were adopting a new piece of technology that demanded one person to operate multiple looms at the same time, making a big leap in production. Many companies closed, while some succeeded and continued thriving. Eliot D. Smith concluded his study with the following conclusion: the companies who survived weren’t just the ones with the best tech (best tech for that era – my comment). They were the ones whose managers talked to workers; change wasn’t imposed but discussed; there was trust and collaboration.
Fast forward, why am I bringing up this century-old story? Because the same principle still applies.
For the last few decades, governments have approached digital transformation as a sequence of technology procurements, information architecture decisions, and modernization roadmaps. But the reality is that technology has never been the problem. Some notable digital projects failures such as Obama Care, Phoenix Pay and Arrive CAN, Heathrow Terminal 5, are examples to reflect on. In the post-mortem analysis of these projects, they did not fail due to bad code alone. They all had serious change management issues. They collapsed because the people, behaviors, incentives, culture, and identity that shape how technology is used were ignored.
ObamaCare, Phoenix, ArriveCan, HT5
Underestimating complexity/scope
Poor planning & testing before full roll-out
Poor stakeholder readiness and engagement
Week oversight, governance, documentation
Emphasis on speed & budget over quality and robustness.
VS
Basic CM principles
Understanding the scale, impact & interdependences of change.
CM includes planning, piloting, & testing new systems with real users.
Managing change requires preparing all stakeholders to accept & adapt to the change.
CM involves clear governance, accountability & documentation to track decisions, responsibilities, & risks.
Effective CM balances timelines, cost & quality.
The tech went forward; the people stayed behind.
Moving on. ..the world after the Covid. According to the International Labour 2022/23 Report, COVID had a structural impact on society and economy equivalent to World War 2, and 4x greater than the financial crisis of 2008. Also,
• 90% of companies indicated that they are going to accelerate digitization.
• 42% of Covid induced permanent layoffs, and,
• 40% of core skills will change for all workers by 2025.
And furthermore, World Economic Forum 2020/21 reported that 84% of companies accelerated digitalization and 50% accelerated automation.
In a post-Covid, interconnected world, digital ambition can no longer be achieved through technical gadgets, tools and systems. It requires a transformation of how people work with those tools, if they believe in those tools, and how they experience the change. Digital transformation has never been about technology. There are multiple strategies across governments from Open Government to zero-copy, to AI directives, and many others. In practice, frontline service delivery still depends on human work (teachers, police, nurses), manual workflows, legacy systems that are not designed to ‘talk’ to each other, siloed program data, Excel workarounds, and processes designed for internal compliance rather than citizen experience.
Every global study by the “Big Four” consulting firms, Deloitte, EY, KPMG and PwC, consistently finds that roughly 60 percent or more of digital, AI, and automation projects fail because of human-centred factors, not technical ones. Employees don’t trust the tools, workflows are not redesigned, leaders fail to effectively communicate change. This causes confusion, and lack of understanding for the why or how of change; people fear replacement, loss of control, or increased oversight.
In a recent article, “A Strategic Approach for Change Managers in Challenging Times”, April K. Mills describes two approaches. One is ‘driving people’ which underscores compliance, mandatory training, fixed deadlines, communications written like policy statements, and performance pressure rather than empowerment. The other is ‘driving change’ which underscores collaborative problem-solving and shared purpose rather than creating barriers, demonstrating personal benefit, and leadership modelling the behavior they expect to see from their employees.
Are we defaulting to the first approach? Not because public servants lack empathy or insight, but due to the bureaucratic nature of many governmental bodies. This can create a disconnect between those making decisions and those affected by those decisions, resulting in a lack of adaptability and responsiveness to the needs of citizens and employees. A waterfall project approach creates risks primarily because it separates the policy and implementation teams. The result is that governments keep launching digital and data initiatives that look strong on paper but collapse upon meeting real human behavior.
Leadership modelling matters.
Digital and AI are changing work in deeply human ways, shaping identity, trust, and purpose. Yet the people-focused teams like HR and change management are often left out. When HR and change management are sidelined, there is no one there to assesses trust dynamics and lack of buy-in from staff.
As I mentioned in one of my previous articles, people spend 10 or 20 years or even more honing their skills, so they identify with it and take pride in their applied skills and trusted tools. Suddenly, with new rollout, identity disruption is ignored as people fear losing the parts of their job/identity they value. Reskilling/in-skilling becomes reactive, not strategic. We need narrative explaining augmentation versus replacement, so it helps staff to not jump to conclusion and imagine the worst, filling the void with their interpretations. This begins by involving HR, frontline staff, program experts and ‘translators’1 at the design stage, not at the end, so their insights shape solutions that reflect real workflows.
Transformation should be approached as a behavioral journey, not a procurement exercise. Pilots must test hypotheses about trust, resistance, and acceptable levels of automation, while success metrics should measure adoption, trust, and workflow improvements, not just system functionality. Technology alone does nothing without belief, trust. Digital transformation is not a technology project; it is a people project.
Note1 – in technology context, “translator” is a person who bridges the communication gap between technical experts/data scientists and non-technical stakeholders. The role involves explaining complex technical/data jargons in simple, relatable business terms.
References
Driving Change: A Strategic Approach for Change Managers in Challenging Times
How Behavioral Science Can Improve the Return on AI Investments
HR 'sidelined' in AI workforce transformation | HRD Canada
49% Of C-suite Executives Struggle With Choosing The Right Tasks To Automate, New Beamery Research F
Annual evaluation report 2022–23 | International Labour Organization
Annual Report 2020-2021 | World Economic Forum
Technology & Labor; by Elliott Dunlap Smith | Open Library
Gartner Survey Reveals Only 48% of Digital Initiatives Are Successful
Make sure to share your own thoughts with the author by leaving a comment below
Log in or sign up to continue the conversation